Mean Fractional Error
mfe.Rd
Compute the Mean Fractional Error (MFE) of observed and predicted data.
Arguments
- obs
A
numeric vector
,data.frame
ordata.table
with observation data.- pred
A
numeric vector
,data.frame
ordata.table
with model (predicted) data.- use
An optional
character vector
giving a method for computing MFE in the presence of missing values. This must be eithereverything
orcomplete.obs
.
Details
The function computes the Mean Fractional Error (MFE) according to EPA formulation: $$ \mathrm{MFE} = 100 \cdot \sum_{i=1}^{N} \left( \frac{|P_i - O_i|}{0.5(P_i + O_i)} \right) $$
The MFE thus normalizes the error for each model-observed pair by the average of the model and observation before taking the average.